The aeration process is often the single largest consumer of electricity in a wastewater treatment plant. Aeration in biological reactors provides microorganisms with oxygen which is required to convert ammonium to nitrate. Ammonium is toxic for aqueous ecosystems and contributes to eutrophication. The importance of aeration for the treatment results in combination with the high costs motivates automatic control of the aeration process. This thesis is devoted to ammonium feedback control in municipal wastewater treatment plants. With ammonium feedback control, the aeration intensity is changed based on a measurement of the outlet ammonium concentration. The main focus of the thesis is design, implementation, evaluation and improvement of ammonium PI (proportional-integral) controllers. The benefits of ammonium feedback control are established through long-term experiments at three large wastewater treatment plants in Stockholm, Sweden. With ammonium feedback control, energy savings up to around 10 % were achieved compared to keeping the dissolved oxygen concentration constant. The experiments generated several lessons learned with regard to implementation and evaluation of controllers in full-scale operation. The thesis has established guidelines on how to design ammonium feedback controllers for situations when cost-effective operation is the overall aim. Simulations have demonstrated the importance to limit the dissolved oxygen concentration in the process and under what conditions the energy saving with ammonium feedback control is large. The final part of the thesis treats improvements of ammonium PI control through minor modifications to the control structure or controller. Three strategies were studied: gain scheduling control, repetitive control, and a strategy reacting to oxygen peaks in the last aerobic zone. The strategies all had their benefits but the ammonium feedback controller was the key factor to improved aeration control.

Cold-formed steel structures have been in service for many years and are used as shelters for both domestic and industrial purposes. To produce an economical product, manufacturers have typically based their designs on the simple portal frame concept. As there is almost a direct relationship between overall cost and the weight of steel in a portal frame structure, it is of great importance to provide a structure with the minimum amount of steel whilst providing structural adequacy. Portal frame sheds have been refined continuously for many years, with only minimal amounts of savings in steel. Therefore, to provide even greater savings in steel, an innovative building system is required. Modern Garages Australia (MGA) is one of the leading cold-formed steel shed manufacturers in Queensland. MGA has recently developed such an innovative building system that has significant economic savings when compared with portal frame structures. The MGA building system has two key differences to that of the conventional portal frame system. These differences are that the MGA system has no conventional frames or framing system, and it has no purlins or girts. This results in the MGA system being completely fabricated from thin cladding, which significantly reduces the quantity of steel. However, the key problem with this building system is that the load paths and structural behaviour are unknown, and therefore the structure cannot be analysed using conventional methods. Therefore, the objectives of this research were to first investigate the structural behaviour of this new building system and its adequacy for an ultimate design wind speed of 41 m/s using full scale testing. The next objectives were to use finite element analysis to optimise the original MGA building system so that it is adequate for an ultimate design wind speed of 41 m/s, and to develop a new improved cold-formed steel building system that has greater structural efficiency than the original MGA building system. This thesis presents the details of the innovative MGA building system, full scale test setup, testing program, finite element analysis of the MGA building system and the results. Details and results from the optimisation of the MGA building system, and the development of a new improved cold-formed steel building system are also presented. The full scale experimental investigation considered the required loadings of cross wind, longitudinal wind and live load test cases and simulated them on the test structure accurately using an innovative load simulation system. The wind loads were calculated for a 41 m/s ultimate design wind speed. Full scale test program included both non-destructive and destructive tests. The finite element analyses contained in this thesis have considered cross wind, longitudinal wind and live load cases, as well as the destructive load case of the MGA building system. A number of different model types were created and their results were compared with the experimental results. In general, two main model types were created. The first type consisted of a 'strip' of the MGA building system (Strip model) and the second modelled the full structure (Full model). Both of these model types were further divided into models which contained no contact surfaces and those which contained contact surfaces to simulate the interfaces between the various components such as the brackets and cladding. The experimental test results showed that the MGA test structure is not suitable for an ultimate design wind speed of 41 m/s. This conclusion is a result of a number of observed failures that occurred during the extensive testing program. These failures included local buckling, crushing failures, and distortional buckling of the cladding panels. Extremely large deflections were also observed. It was calculated that for the MGA building system to be adequate for the design wind speed of 41 m/s, a cladding thickness of 0.8 mm was required. This also agreed well with the finite element analysis results which concluded that a cladding thickness of 0.8 mm was required. In order to avoid the increased use of steel in the building system, a new improved cold-formed steel building system was developed and its details are provided in this thesis. A finite element model of this new improved cold-formed steel building system was created and the results showed that the new building system was able to achieve a load step equivalent to an ultimate design wind speed of 50.4 m/s and was approximately 250% stiffer than the original MGA building system, without any increase in the overall weight of the building system. It is recommended that this new improved cold-formed steel building system be further developed with the aid of finite element modelling and be tested using a similar full scale testing program that was used for the original MGA building system.

Thesis (MScEng)--Stellenbosch University, 2013. / Wind energy has proven to be a viable source of clean energy, and the worldwide demand is growing rapidly. Variable speed topologies, with synchronous generators and full-scale converters, are becoming more popular, and the e ective control of these systems is a current trend in wind energy research. The purpose of this study is the modelling, design, simulation and implementation of a small-scale, variable speed wind energy conversion system, incorporating the position sensorless direct torque control with space vector modulation, of a permanent magnet synchronous generator, including an RLC converter lter. Another aim is the development of a gain scheduling algorithm that facilitates the high level control of the system. Mathematical models of the combined lter-generator model, in the stationary and rotating reference frames, are presented and discussed, from which equivalent approximate transfer functions are derived for the design of the controller gains. The design of the controller gains, RLC lter components, gain scheduling concept and maximum power point tracking controller are presented. It is discovered that the RLC lter damping resistance has a signi cant e ect on the resonance frequency of the system. The system is simulated dynamically in both Simulink and the VHDL-AMS programming language. Additionally, the maximum power point tracking controller is simulated in the VHDL-AMS simulation, including a wind turbine simulator. The simulation results demonstrate good dynamic performance, as well as the variable speed operation of the system. The practical results of torque and speed controllers show satisfactory performance, and correlate well with simulated results. The detailed gain scheduling algorithm is presented and discussed. A nal test of the complete system yields satisfactory practical results, and con rms that the objectives of this thesis have been reached.

La ventilation est un secteur clé du bâtiment, dont le rôle est d’assurer un air sain et confortable toute l’année, ce qui est rarement le cas dans les bâtiments scolaires, tout en minimisant les consommations énergétiques. Nos travaux consistent à apporter des éléments de réponses et des pistes d’amélioration pour l’élaboration de stratégies de ventilation appropriées au milieu scolaire, à partir de travaux expérimentaux et numériques. Pour ce faire, quatre modes de ventilation (naturelle et mixte) ont été testés dans des salles de classes de l’Université de La Rochelle, et leurs performances ont été comparées via une évaluation multicritère basée sur les paramètres physiques caractérisant l’environnement intérieur, les indices de confort (subjectif, analytique et adaptatif) et des critères énergétiques. Des méthodes normatives de classification et des estimations de consommations énergétiques nous ont permis de conclure, que le système de ventilation naturelle par ouverture automatisée des fenêtres, contrôlé par la détection de présence et des paramètres thermiques (système SOS), est le meilleur compromis. Nous soumettons toutefois dans ce manuscrit, un certain nombre d’améliorations à apporter à ce système.Notre contribution porte également sur la connaissance des mécanismes du confort humain et en particulier ses réactions adaptatives, en définissant les conditions favorables au confort et en proposant des modèles prédictifs du confort global, de l’ajustement personnel ainsi que du contrôle individuel de l’ambiance par les occupants. Ces résultats ont notamment pour vocation d’améliorer la prise en compte des interactions entre les occupants et leur environnement dans les simulations numériques et pourraient également servir de base au développement d’une stratégie de ventilation optimisée. Au niveau numérique, nous proposons des simulations annuelles de quatre stratégies de ventilation, très proches de celles testées sur site, à l’aide d’un code thermo-aéraulique multizone (couplage Trnsys/Contam), que nous avons validé à partir de certaines séquences de mesures. En tenant compte d’un scénario d’occupation scolaire standard et du fichier météorologique correspondant à la station de La Rochelle, nous avons notamment montré qu’il est primordial de pré-chauffer l’air d’un système de ventilation mécanique, sous peine d’être confronté à des besoins de chauffage insurmontables. En introduisant une puissance de chauffage illimitée, permettant de maintenir une température minimale acceptable et ainsi de simuler des conditions d’enseignement réalistes, il apparaît que la meilleure qualité environnementale intérieure est à nouveau obtenue avec le système SOS. Notre modèle nous donne désormais la possibilité de multiplier les stratégies de ventilation, ainsi que les scénarios d’occupation, les conditions climatiques ou tout autre étude paramétrique, afin d’élaborer les meilleures stratégies de ventilation dans chaque configuration. / Ventilation is a key sector of building, whose role is to ensure healthy and comfortable air all over the year, which is rarely the case in school buildings, while minimizing energy consumption. Our work provides some answers and possible improvements for the development of appropriate ventilation strategies for schools, from experimental and numerical work.To achieve this, four modes of ventilation (natural and mixed ventilation modes) were tested in classrooms of the University of La Rochelle, and their performances were compared via a multicriteria evaluation based on the physical parameters characterizing the indoor environment, comfort indices (subjective, analytical and adaptive) and energy criteria. Normative methods of classification and estimates of energy consumption enabled us to conclude that the natural ventilation system by automated opening windows, controlled by the presence detection and thermal parameters (SOS), is the best compromise. However, we submit in this manuscript, some improvements to this system.Our contribution concerns also the understanding of the human comfort mechanisms and in particular its adaptive reactions, by defining the favorable conditions for a state of comfort and providing predictive models concerning overall comfort, personal adjustments and the individual control of the indoor environment by the occupants. These results aim to improve the consideration of the interactions between occupants and their environment in numerical simulations, and may serve as a basis for developing an optimized ventilation strategy.Numerically, we propose annual simulations of four ventilation strategies, very similar to those tested in situ, by using a combined heat and mass transfer multizone model (coupling Trnsys / CONTAM), that we have validated from selected experimental sequences. Taking into account a standard scenario of occupation and the annual weather conditions for La Rochelle, we have shown the importance to pre-heat the supplied air of a mechanical ventilation system, because of insurmountable heating demand consequences. By introducing an unlimited heating power, in order to maintain a minimum acceptable temperature and thus to simulate realistic learning conditions, it appears that the best indoor environmental quality is again obtained with the SOS system. Our model now gives us the possibility to increase the number of ventilation strategies, as well as the occupation scenarios, the weather conditions or any other parametric study in order to design the best ventilation strategies for each configuration.

Rubber asphalt has been used in Brazil as early as 2001. Among the fabrication processes of rubber asphalt, the most widely used in Brazil is the terminal blend. However, the use of field-blended rubber asphalt has been around in the United States for decades, especially in the state of Arizona. This process results in a highly viscous material with enhanced engineering properties, but requires specific equipment that is typically installed at the job site, or close to the supplying asphalt plant. Thus, keeping in mind the possible technological advantages of using a field blended rubber asphalt mixture and the lack of information regarding this technique in Brazil, there is a necessity to develop studies to assess the performance of this type of material in our environmental conditions and axle loading configurations. Therefore, a laboratory study was conducted to determine the rheological properties of a field-blended rubber asphalt and compare them to those of typical binders used in Brazil (an AC 30/45 penetration grade and a binder modified with SBS, an elastomeric polymer). Binder permanent deformation was determined using the Multiple Stress Creep and Recovery (MSCR) test, whereas binder fatigue behavior was determined using the Time Sweep and Linear Amplitude Sweep (LAS) tests. Subsequently, the permanent deformation and fatigue behavior of a gap-graded mixture using the field-blended rubber asphalt were assessed in the laboratory and in the field. The permanent deformation of the mixture was determined in the laboratory using the LCPC wheel track test, whereas the fatigue behavior was determined using the four point bending flexural test. A test section was built after the rehabilitation job of highway RJ-122, where a field-blended rubber asphalt mixture was first used in the country. The mixture performance was studied in situ through accelerated pavement tests using a full scale, large mobile traffic simulator. The results were used to model the performance of the structure with the rubber asphalt mixture by means of the Highway Development and Management Model (HDM-4) cracking and permanent deformation models, calibrated to local conditions. Accelerated pavement tests were validated through periodic pavement monitoring campaigns conducted for four years in a test section in Highway RJ-122. The field-blended rubber asphalt showed a good performance in terms of permanent deformation and fatigue determined in the laboratory at the binder and mixture levels, which confirms what was verified in the field. / No Brasil, o asfalto-borracha vem sendo utilizado desde meados de 2001. Dentre os processos de fabricação do asfalto-borracha, o mais utilizado no Brasil é o asfalto-borracha estocável ou terminal blend. Contudo, o asfalto-borracha do tipo não estocável (field blend) vem sendo bastante difundido nos Estados Unidos há décadas, principalmente no estado do Arizona. Este processo resulta em um asfalto-borracha de alta viscosidade, com alto desempenho, porém requer um equipamento de fabricação de asfalto-borracha específico, instalado no canteiro de obras, ou muito próximo à usina de asfaltos fornecedora da obra. Tendo em vista as possíveis vantagens tecnológicas do asfalto-borracha field blend e o conhecimento ainda pequeno sobre esta técnica no Brasil, há a necessidade de desenvolver estudos para a avaliação deste tipo de material frente às condições climáticas e de solicitação pelas cargas viárias em nosso país. Para tanto, foi realizado um estudo em laboratório para verificar as propriedades reológicas de um asfalto-borracha field blend e compará-las com as características de ligantes típicos empregados no Brasil (um CAP 30-45 e um ligante modificado por polímero elastomérico do tipo SBS). Elegeu-se o Multiple Stress Creep and Recovery (MSCR) para verificar a deformação permanente e o Time Sweep e Linear Amplitude Sweep (LAS) para verificar o comportamento na fadiga dos ligantes. Na sequência, determinou-se a deformação permanente e o comportamento à fadiga de uma mistura asfáltica descontínua (gap-graded) empregando o asfalto-borracha field blend em laboratório e no campo. A deformação permanente da mistura foi verificada por meio do simulador de tráfego LCPC, enquanto a vida de fadiga foi determinada utilizando o ensaio de flexão em viga (4 pontos). Por fim, foi construída uma seção teste após a conclusão da obra de restauração dos pavimentos da rodovia RJ-122, local onde se elegeu a utilização desta tecnologia pela primeira vez no país. O desempenho da mistura foi estudado in loco com o emprego de ensaios acelerados do pavimento utilizando-se o simulador de tráfego linear móvel em tamanho real. Os resultados obtidos foram utilizados para modelar o desempenho da estrutura com o revestimento asfáltico constituído pela mistura com o asfalto-borracha por meio dos modelos de trincamento e de deformação permanente do Highway Development and Management Model (HDM-4), podendo-se verificar o desempenho desta mistura calibrado para as condições locais. Os ensaios acelerados foram validados em campo por meio de campanhas de monitoramento periódicas realizadas ao longo de quatro anos na rodovia RJ-122. Pelos ensaios de laboratório no ligante e na mistura foi possível concluir que o asfalto-borracha field blend apresenta um bom desempenho quanto à deformação permanente e à fadiga, corroborando o que foi verificado no campo.

Le processus de digestion anaérobie conduit à la production de biométhane, un vecteur flexible d’énergie renouvelable. L’amélioration du rendement de ce processus est souvent évoquée comme dépendante de la compréhension approfondie de la structure et de la dynamique des communautés microbiennes qui y sont impliquées. L’objectif de la thèse a été de caractériser les communautés microbiennes impliquées dans le processus de digestion anaérobie et de déterminer l’influence de facteurs opérationnels sur leurs dynamiques. Nous nous sommes en particulier intéressés à l’augmentation du taux de charge organique, le type de digesteurs anaérobies (réacteur continu perpétuellement mélangé vs réacteur anaérobie à chicane), mais aussi à la phase de démarrage d’un digesteur de ferme avec une montée en température. En absence de conditions contraignantes, nous avons observé l’installation de populations méthanogènes les mieux adaptées à la production de biogaz dans les réacteurs étudiés et la mise en place de communautés microbiennes similaires entre réacteurs réplicats. Cependant, des changements au niveau opérationnel ont conduit au développement de communautés divergentes en termes de structure. En effet, en présence d’un environnement déterministe, la plupart des bactéries et archées impliquées en digestion anaérobie ont montré une redondance fonctionnelle à la perturbation. Toutefois, certaines populations bactériennes dominantes ont également pu montrer des phénomènes de résistance, en termes de présence et d’abondance, à l’évolution des conditions environnementales. Au cours de nos études, les différentes communautés s’installant dans les digesteurs étudiés ont également montré des aptitudes variables pour la production de biogaz. De plus, des corrélations entre les communautés bactériennes, archées et eucaryotes ont aussi été démontrées, soulignant le rôle non négligeable des eucaryotes dans le processus de digestion anaérobie et l’installation de communautés microbiennes dominantes et spécifiques à la production de biogaz. Ainsi, les changements au sein de la communauté microbienne résultant de la modification progressive de facteurs opérationnels, et ce bien avant l’apparition des premiers symptômes d’inhibition de la production de biogaz, pourraient permettre le développement d’indicateurs microbiens de l’état du processus de digestion anaérobie et donc la mise en place d’une gestion microbiologique raisonnée des digesteurs anaérobies / The anaerobic digestion process leads to the production of biomethane, a versatile renewable energy vector. The dynamics and interactions between specific microbial groups are currently considered as key research subjects towards the improvement of the anaerobic digestion (AD) process. Indeed, deeper knowledge of the ecology of AD, the dynamics of the microbial populations and their structure could provide valuable information regarding unexplained and unpredictable failures or malfunctioning of the anaerobic digestion process. The aim of this work was to characterize the microbial communities involved in the AD process, and to study their responses due to the change of operational parameters such as an increase of the organic loading rate, the reactor type (completely stirred tank reactor vs anaerobic baffled reactor), or the start-up phase of a farm reactor with a shift from psychrophilic to mesophilic temperature range. While we observed the installation of similar microbial populations between replicated reactors under stable conditions, best adapted to biogas production, the microbial communities started to diverge once the operational parameters changed. Indeed, due to deterministic environment, most of bacteria and archaea showed redundant functional adaptation to the changing environmental conditions. However, some dominant bacterial populations were also resistant in terms of presence and abundance to the environmental change. The specific microbial communities established in our studied reactors showed also discrepancies in terms of biogas yields. Furthermore, correlations between the bacterial, archaeal and eukaryotic communities were pointed out, indicating the putative influence of eukaryotes on the anaerobic digestion process and the establishment of the other microbes having crucial functions during the anaerobic biomass digestion. Interestingly, shifts inside the anaerobic microbial community due to the gradual change of operational parameters, were detected prior to any biogas production inhibition, giving the opportunity for the development of potential early microbial indicators for assessing the AD process status and improving the microbial management of anaerobic reactors

Pour les poutres et les dalles ne comportant pas d'armatures de cisaillement, la résistance à l'effort tranchant ou au poinçonnement est souvent un critère important de dimensionnement. Ce type de rupture est caractérisé par un comportement fragile pouvant conduire à l'effondrement partiel voir total de la structure. Malgré de nombreuse recherche dans ce domaine, la résistance à l'effort tranchant et au poinçonnement des structure en béton armé ou précontraint demeure un phénomène complexe et dont l'approche normative est souvent empirique est simplifiée. La capacité des bétons renforcés de fibres métalliques à réduire voir à remplacer totalement les armatures de cisaillement des structures en béton armé et précontraint a été mis en évidence par plusieurs études expérimentales. Cependant, et malgré ses nombreux atouts, l'application à l'échelle industrielle des bétons de fibres est restée marginal, principalement due au manques d'un cadre normatif cohérent et reconnu. Les processus fixes d'une usine de préfabrication d'éléments en béton offre des possibilités optimales pour utiliser des matériaux cimentaires à hautes performances tel que les bétons autoplaçant, les bétons à hautes résistances, etc. Du point de vue de l'auteur, l'utilisation de bétons à hautes performances renforcés de fibres métalliques est le pas de développement et d'optimisation pour cette industrie. Les Bétons Fibrés à Hautes Performances (BFHP) reprennent une matrice similaire aux Bétons à Hautes Performances (BHP) auxquels est ajouté une certaine quantité de fibres métalliques conférant au matériau un comportement au niveau de la structure exploitable dans le dimensionnement. Les BFHP présentent un ratio résistances/coûts intéressant ainsi qu'une alternative au Béton Fibré Ultra-Performants (BFUP). L'objectif principal de ce travail est d'analyser le comportement au cisaillement et au poinçonnement d'éléments de structures en BFHP et en BFUP sans armatures de cisaillement et proposé des recommandations et des règles de dimensionnement adaptées aux ingénieurs de la pratique (…) / For members and flat slabs without shear reinforcement, the shear and punching shear strength are often the determining design criteria. These failure modes are characterized by a fragile behaviour implying possible partial or total collapse of the structure. Despite extensive research in this field, shear and punching shear in reinforced and prestressed concrete structures, remain complex phenomena so much that the current approach is often empirical or simplified. The ability of Steel Fibre Reinforced Concrete (SFRC) to reduce shear reinforcement in reinforced and prestressed concrete members and slabs,or even eliminate it, is supported by several experimental studies. However its practical application remains marginal mainly due to the lack of standard, procedures and rules adapted to its performance. The stationary processes in precast industry offer optimal possibilities for using high performance cementitious materials such as Self Compacting Concrete (SCC) and High Strength Concrete (HSC). For the author, the combination of High Performance Concrete and steel fibres is the following step in the development and the optimization of this industry. The High Performance Fibre Reinforced Concrete (HPFRC) stands between conventional SFRC and Ultra-High Performance Fibre Reinforced Concrete (UHPFRC). The HPFRC exhibiting a good strength/cost ratio is, thus, an alternative of UHPFRC for precast elements. The principal aim of this work was to analyse the shear and punching shear behaviour of HPFRC and UHPFRC structures without transversal reinforcement and to propose recommendations and design models adapted for practitioners. Several experimental studies on structural elements, i.e. beams and slabs, were undertaken for this purpose. Firstly, an original experimental campaign was performed on pre-tensioned members in HPFRC. A total number of six shear-critical beams of a 3.6 m span each, and two full scale beams of a 12 m span each, were tested in order to evaluate the shear and flexural strength. The principal parameter between the specimens was the fibres (…)

Les voies ferrées, construites pour la plupart depuis plus d'une centaine d'années, sont des ouvrages vieillissants. Elles nécessitent une maintenance et un entretien accrus, ce qui constitue un enjeu technique et économique majeur pour les années à venir. Jusqu'à l'ouverture des marchés à la concurrence, la mise en œuvre des voies nouvelles était vérifiée empiriquement par la SNCF. Du fait de la mise en œuvre de la directive européenne 91/440/10, l'exploitant historique se tourne d'un objectif de moyen vers un objectif de résultat. Cela nécessite donc de disposer de méthodes d'auscultation non destructives, permettant de vérifier que le compactage du ballast est correctement réalisé, avant de faire circuler le trafic voyageurs, en vue de garantir un niveau de sécurité élevé. Cette première approche pourrait être poursuivie pour assurer une auscultation à grand rendement. Cependant, les méthodes existantes permettant d'obtenir l'état de compactage du ballast à la mise en œuvre sont ponctuelles et difficiles à mettre en place; elles ne répondent pas à la problématique de doublement de la maintenance des voies des prochaines années. L'étude de la propagation d'ondes vibratoires dans le ballast est une alternative à ces méthodes qui peut permettre de répondre à ces contraintes. Le ballast est un milieu discontinu complexe pour la compréhension des ondes car elles se propagent dans un chaînon de force. Il présente une grande difficulté dans la modélisation du fait de la taille élevée des éléments et doit être traité comme un milieu discret ne répondant pas à une mécanique élastique de milieu continu. Étant donné la difficulté de modéliser cette couche discrète, il convient de traiter le problème par l'expérimentation. L'objectif de cette thèse est donc d'orienter la recherche vers l'utilisation de la propagation des ondes vibratoires dans la structure de la voie. Ce mémoire est organisé comme suit :- un premier chapitre détaille la structure de la voie ferrée et le matériau granulaire qu'est le ballast, ainsi que les méthodes de diagnostic des voies ferrées existantes.- le deuxième chapitre décrit les différents types d'ondes vibratoires se propageant dans un milieu élastique homogène, puis dans le ballast, et étudie la réponse du ballast à travers celle de la traverse ferroviaire.- Ces deux chapitres, issus de l'état de l'art, permettent de définir dans le chapitre trois les expérimentations réalisées dans le cadre de ce travail sur une structure ferroviaire en vraie grandeur : mise en œuvre, instrumentation, résultats. Ce chapitre s'attache particulièrement à décrire la vitesse des ondes et leur amortissement dans le ballast, les courbes de dispersion mesurées.- Enfin, la propagation d'une onde vibratoire dans le ballast est étudiée dans le chapitre quatre par le biais d'une simulation numérique, avec la comparaison de deux modélisations discrète et continue avec l'expérimentation / Railways, most of them built for over one hundred years, are old structures. They require increasing maintenance, a major technical and economic challenge for the coming years. Until the opening of markets to competition, the implementation of new railroads was empirically controlled by SNCF. Because of the application of the European directive 91/440/10, the historical operator turns from a goal of means to a goal of results. This calls for non-destructive highly efficient auscultation methods to check the right compaction of the ballast. However, existing methods for obtain ballast compaction assessment during implementation stage are surface-limited and difficult to apply; they do not respond to the issue of the doubling of track maintenance. Thus the study of the propagation waves in the ballast is an alternative to these methods and may allow to answer these requirements. The ballast is a discontinuous medium, complex for the understanding of waves, as they propagate in a force link. Modelling ballast is especially difficult because of the large size of its components and because it should be treated as a discrete environment, not following the elastic mechanics of continuous media. Given the difficulty to model this discrete layer, it is necessary to undertake the problem with experimentation. The aim of this thesis is to focus on the use of the propagation of vibration waves in the railroad structure. This work is organized as follows : the first chapter details the structure of the railroad and the ballast as a granular material, and the existing assessment methods for railroad. The second chapter describes the different types of vibration waves that propagate, first, in an elastic homogeneous medium, second in the ballast, and presents the answer of ballast through the answer of the sleeper. The two previous chapters, derived from the state of the art, allow to define in chapter three the experiments implemented as part of this work on a full scale railroad structure : realization, instrumentation, results. This chapter especially endeavours to describe the waves celerity, their damping in ballast and the measured dispersion curves. Finally, the propagation of a vibration wave in the ballast is studied in chapter four and a numerical simulation, compares with a disctete model and a continuous model with experimental results

Widespread damage to roofing materials (such as tiles and shingles) for low-rise buildings, even for weaker hurricanes, has raised concerns regarding design load provisions and construction practices. Currently the building codes used for designing low-rise building roofs are mainly based on testing results from building models which generally do not simulate the architectural features of roofing materials that may significantly influence the wind-induced pressures. Full-scale experimentation was conducted under high winds to investigate the effects of architectural details of high profile roof tiles and asphalt shingles on net pressures that are often responsible for damage to these roofing materials. Effects on the vulnerability of roofing materials were also studied. Different roof models with bare, tiled, and shingled roof decks were tested. Pressures acting on both top and bottom surfaces of the roofing materials were measured to understand their effects on the net uplift loading. The area-averaged peak pressure coefficients obtained from bare, tiled, and shingled roof decks were compared. In addition, a set of wind tunnel tests on a tiled roof deck model were conducted to verify the effects of tiles’ cavity internal pressure. Both the full-scale and the wind tunnel test results showed that underside pressure of a roof tile could either aggravate or alleviate wind uplift on the tile based on its orientation on the roof with respect to the wind angle of attack. For shingles, the underside pressure could aggravate wind uplift if the shingle is located near the center of the roof deck. Bare deck modeling to estimate design wind uplift on shingled decks may be acceptable for most locations but not for field locations; it could underestimate the uplift on shingles by 30-60%. In addition, some initial quantification of the effects of roofing materials on wind uplift was performed by studying the wind uplift load ratio for tiled versus bare deck and shingled versus bare deck. Vulnerability curves, with and without considering the effects of tiles’ cavity internal pressure, showed significant differences. Aerodynamic load provisions for low-rise buildings’ roofs and their vulnerability can thus be more accurately evaluated by considering the effects of the roofing materials.

Hurricane is one of the most destructive and costly natural hazard to the built environment and its impact on low-rise buildings, particularity, is beyond acceptable. The major objective of this research was to perform a parametric evaluation of internal pressure (IP) for wind-resistant design of low-rise buildings and wind-driven natural ventilation applications. For this purpose, a multi-scale experimental, i.e. full-scale at Wall of Wind (WoW) and small-scale at Boundary Layer Wind Tunnel (BLWT), and a Computational Fluid Dynamics (CFD) approach was adopted. This provided new capability to assess wind pressures realistically on internal volumes ranging from small spaces formed between roof tiles and its deck to attic to room partitions. Effects of sudden breaching, existing dominant openings on building envelopes as well as compartmentalization of building interior on the IP were systematically investigated. Results of this research indicated: (i) for sudden breaching of dominant openings, the transient overshooting response was lower than the subsequent steady state peak IP and internal volume correction for low-wind-speed testing facilities was necessary. For example a building without volume correction experienced a response four times faster and exhibited 30-40% lower mean and peak IP; (ii) for existing openings, vent openings uniformly distributed along the roof alleviated, whereas one sided openings aggravated the IP; (iii) larger dominant openings exhibited a higher IP on the building envelope, and an off-center opening on the wall exhibited (30-40%) higher IP than center located openings; (iv) compartmentalization amplified the intensity of IP and; (v) significant underneath pressure was measured for field tiles, warranting its consideration during net pressure evaluations. The study aimed at wind driven natural ventilation indicated: (i) the IP due to cross ventilation was 1.5 to 2.5 times higher for Ainlet/Aoutlet>1 compared to cases where Ainlet/AoutletCFD based IP responses. Comparisons with ASCE 7-10 consistently demonstrated that the code underestimated peak positive and suction IP.